Visual data, which means data can be displayed and seen by a user (ex. image data or video data), always needs to be further processed to meet different requirements.
For examples, the visual data may need to be compressed to save transmission bandwidth or storage space. The de-compressed visual data corresponding to different compression algorithms have different qualities. For more detail, in one example, if decompressed visual data with high quality is desired, the compression algorithm should comprise few steps, less compressing time, or the decompressed visual data may have distortion. On the contrary, if a compression algorithm with complicated steps, and long compressing time is applied, the corresponding decompressed visual data may have poor quality since more distortion for decompressed visual data may exist. Also, compression ratios for different compression algorithms are also different. Therefore, it is hard to select a proper compression algorithm for visual data since the visual data is always compressed via a single compression algorithm.